Citadel expands AI talent hunt as hedge funds compete with technology firms

Citadel plans to expand its quantitative investment team at a healthy double-digit rate over the next year, as hedge funds increasingly compete with artificial intelligence laboratories and technology companies for specialist researchers.

According to Bloomberg, Navneet Arora, Citadel's global head of quantitative strategies, said the group has about 180 financial researchers and data engineers and will continue recruiting university graduates alongside experienced AI researchers.

The hedge fund is also approaching researchers at AI laboratories including Google DeepMind, seeking specialists capable of generating investment ideas and applying increasingly powerful AI systems to trading.

AI reshapes quantitative research

Arora said the growing use of AI has not eliminated the need for human researchers. Instead, he said, it is changing the skills that quantitative investment firms value.

The increasing automation of tasks such as data collection and analysis is reducing the importance of programming expertise, according to Arora, while increasing the value of researchers who can develop investment ideas and direct automated systems.

“Now one person can effectively manage two, four, and in some cases even more processes. Humans are becoming ‘managers of machines’,” Arora said.

“And these people are not as easily replaced as machines.”

Citadel's website indicates that a new team led by Alexey Poyarkov has also been established to focus on systematic global equity trading. Poyarkov recently joined Citadel from TGS Management.

Expansion across strategies

Citadel is seeking to use its expanded workforce to strengthen its core equity business and develop strategies in areas including futures and volatility.

The firm manages about US$76 billion in assets across equities, fixed income and commodities, with billions of dollars currently allocated to quantitative strategies.

Citing source familiar with the matter, Bloomberg reported that Citadel's Tactical Trading fund, which combines fundamental stock investing with quantitative strategies, gained 24.7% year-to-date through August.

The same source said the fund has achieved an annualised return of about 20% since its inception in 2008.

The wider talent competition

The competition for AI specialists extends beyond traditional rivals such as banks, hedge funds and proprietary trading firms.

Quantitative investment firms and trading companies are competing for machine-learning researchers and engineers who are also sought by AI laboratories and major technology companies.

Firms including Citadel, Two Sigma, D. E. Shaw, Renaissance, Millennium, Point72 and Balyasny, alongside proprietary trading firms such as Jane Street, Hudson River Trading and XTX, have been reported to offer substantial compensation and research environments to attract specialist talent.

Some have also established dedicated AI-driven strategies or units.

The competition operates in both directions, with AI laboratories also recruiting quantitative researchers and engineers from the financial sector.

Gerald Beeson, chief operating officer at Castle Investments, said the development was creating new opportunities for recruitment.

“This opens up a new frontier for our hiring. Solving complex, real-world business problems is very appealing to the talent we're looking to recruit,” Beeson said.

AI and the durability of trading signals

The rapid adoption of AI is also raising questions about how long profitable quantitative trading strategies can retain their advantages as more investors gain access to similar technologies.

A recent paper by researchers at New York University found that the excess return of a profitable trading signal can halve in about 18 months. Before widespread AI adoption, the process took five to seven years.

Arora said the impact of machine learning varies considerably between trading areas.

In short-term trading, machines can process large quantities of volume and price-movement data.

However, Arora said the advantages in these areas are relatively limited and easier for competitors to replicate, potentially causing excess returns to disappear more quickly.

Longer-term strategies depend on scarcer data and more complex models to generate trading signals.

Arora said this could make their advantages harder for competitors and machines to identify, potentially allowing them to persist for longer.

The shift is consequently changing the competition for quantitative talent, with financial firms seeking researchers who can combine investment judgement with increasingly automated AI systems.